The development of a valid and reliable nutrition knowledge questionnaire and performance-rating scale for urban South African adolescents participating in the 'Birth-to-Twenty' study

THE DEVELOPMENT OF A VALID AND RELIABLE NUTRITION

KNOWLEDGE QUESTIONNAIRE AND PERFORMANCE-RATING

SCALE FOR URBAN SOUTH AFRICAN ADOLESCENTS

PARTICIPATING IN THE ‘BIRTH-TO-TWENTY’ STUDY

by

LINDIWE HARRIET WHATI, B.Sc Dietetics

Thesis presented for the partial fulfilment of the requirements for the degree

of

M.Sc Nutrition Science

at the

UNIVERSITY OF STELLENBOSCH

Study leader:

Dr. M Senekal

Co-study leader: Dr. NP Steyn

Statistical advisors: Dr CJ Lombard and Dr JH Nel

DECLARATION

I, the undersigned hereby declare that the work contained in this thesis is my

own original work and that I have not previously in its entirety or impart

submitted it at any University for a degree.

_______________________ ____________

SUMMARY

The Birth to Twenty (BTT) study involves the monitoring of the health status and related factors of urban-born children from birth until age twenty. When the cohort reached age 13 years in 2003, nutrition knowledge assessment was identified as an important new priority and a nutrition knowledge questionnaire was required for these purposes. Subsequently a valid and reliable nutrition knowledge questionnaire was developed for the BTT study. This process started with the development of a conceptual framework of nutrition-related issues facing urban South African adolescents and identification of related nutrition concepts. A pool of potential questionnaire items reflecting the concepts was subsequently developed. These items were evaluated by an expert panel to ensure content and face validity before being structured into a questionnaire. The resulting 88-item questionnaire was completed by adult and adolescent samples, each age group comprising subgroups of those likely to have good nutrition knowledge and those likely to have poor nutrition knowledge. The data obtained from the completion of the questionnaire by these groups was used to refine the questionnaire through the determination of difficulty and discriminatory indices of the items, and the deletion of items that did not meet the stated criteria. The construct validity of the remaining 63 items was assessed using the same data set. To assess the internal consistency reliability (ICR) of the 63-item questionnaire it was completed by an adolescent sample population considered to be representative of the BTT cohort, after which the questionnaire underwent further steps of refinement. The result was a 60-item questionnaire of which the ICR and construct validity was reassessed and found to be satisfactory. However, to ensure the accurate interpretation of scores obtained by testees, the development of a performance-rating scale was necessary.

A norm-referenced performance-rating scale (norms) was developed by administering the nutrition knowledge questionnaire to a sample population similar to the BTT cohort (norm group) and transforming their performance scores to z-scores. The z-scores ranges were then categorised into stanines, thereby resulting in a norm-referenced performance-rating scale that can be used to rate the performance of the BTT cohort. The validity of the norms was assessed by administering the nutrition knowledge questionnaire to three validation groups that comprised groups who were expected to obtain different performance-ratings on the questionnaire based on their varying levels of nutrition knowledge. The validation groups performed as expected, with significant differences in performance-rating profiles found among the three groups, indicating the validity of the norms.

The study was successful in developing a reliable and valid nutrition knowledge questionnaire for use on the urban adolescents who participate in the BTT study. A norm-referenced performance-rating scale for use with the questionnaire was also successfully developed. The questionnaire and norms will be useful in assessing nutrition knowledge as well as in comparing the changes in knowledge of the BTT cohort as they move from lower to higher school grades.

OPSOMMING

Die “Birth to Twenty” (BTT)-studie behels die monitering van die gesondheidstatus en verwante faktore van kinders wat in stedelike gebiede gebore is van geboorte tot twintigjarige ouderdom. Toe die kohort in 2003 dertienjarige ouderdom bereik het, is die evaluering van die voedingkennis van die kinders as ‘n belangrike nuwe prioriteit geïdentifiseer. ‘n Toepaslike voedingkennisvraelys is vir hierdie doeleindes benodig en gevolglik is ‘n geldige en betroubare vraelys vir gebruik in die BTT-studie ontwikkel. Hierdie proses is begin deur die ontwikkeling van ‘n konseptueleraamwerk oor voedingverwante vraagstukke wat stedelike Suid-Afrikaanse adolessente in die gesig staar, asook die identifisering van verwante voedingkonsepte. ‘n Poel van potensiële vraelysitems wat die konsepte reflekteer is daarna ontwikkel. Die items is eers deur ‘n paneel van kenners evalueer om inhoud- en gesigsgeldigheid te verseker alvorens dit in ‘n vraelys omskep is. Die produk was ‘n vraelys wat 88 items ingesluit het wat vervolgens deur volwasse en adolessente groepe voltooi is. Die groepe het subgroepe ingesluit van diegene met verwagte goeie voedingkennis en diegene met verwagte swak voedingkennis. Die data wat tydens hierdie stap gegenereer is, is gebruik om die vraelys verder te verfyn deur die bepaling van die moeilikheids- en diskriminatoriese-indekse van die items. Die items wat nie aan vooraf gestelde kriteria voldoen het nie, is weggelaat. Die konstrukgeldigheid van die oorblywende 63 items is bepaal deur dieselfde datastel te gebruik. Om die interne-konsekwensie-betroubaarheid (IKB) van die vraelys te bepaal, is dit deur ‘n steekproef van adolessente, wat verteenwoordigend van die BTT-kohort is, voltooi. Hierna is die vraelys verder verfyn. Die uitkoms was ‘n 60-item vraelys waarvan die IBR en konstrukgeldigheid weereens bepaal is. Dit is gevind dat dié twee indikatore van geldigheid en betroubaarheid bevredigend is. Om akkurate interpretasie van die punt wat deur ‘n respondent vir die toets behaal te verseker, is die ideal om ‘n skaal te ontwikkel wat gebruik kan word om dié punt te takseer.

‘n Norm-gebaseerde prestasietakseringskaal is ontwikkel deur die voedingkennisvraelys deur ‘n steekproef wat verteenwoordigend is van die BTT-kohort (normgroep), te laat voltooi. Die prestasiepunte is getransformer na z-tellings wat vervolgens getransformeer is na stanneges, wat ‘n norm-gebaseerde prestasietakseringskaal opgelewer het wat gebruik kan word om die prestasie van die BTT-kohort te takseer. Valideringsgroepe met verskillende vlakke van voedingkennis, wat dus na verwagting verskillend getakseer behoort te word indien die norme toegepas word, het die voedingkennisvraelys voltooi om die geldigheid van die norme te bepaal. Dié valideringsgroepe het

soos voorspel presteer, met betekenisvolle verskille in die prestasieprofiele van die verskillende groepe. Hierdie resultate dui daarop dat die norme geldig is.

Die ontwikkeling van ‘n geldige en betroubare voedingkennistoets vir gebruik in die BTT-studie is suksesvol in hierdie studie deurgevoer. ‘n Norm-gebaseerde prestasietakseringskaal vir gebruik saam met die vraelys is ook suksesvol ontwikkel. Die vraelys en norme sal van waarde wees vir die evaluering van die voedingkennis van die BTT-kohort. Dit sal ook met sukses gebruik kan word om die verandering in die voedingkennis van die kinders soos wat hulle ouer word, te bepaal.

ACKNOWLEDGEMENTS

My sincerest gratitude to Marjanne and Nelia for their total commitment towards my studies and most importantly, believing in me. Dr Lombard and Dr Nel for the expertise and services that they provided when the research was being conducted as well as in the writing up process. To the Medical Research Council, specifically Dr K Steyn, Dr C Nonkwelo and staff of the Chronic Diseases of Lifestyle unit, for granting me the opportunity to work in the environment that is the MRC with all the people that I have come to admire and respect. To Dr S Norris and staff of the ‘Birth-to-Twenty’ study, for the funding and support. To Jean Fourie, Jannetta Cilliers and Theo Nell, for your helping hands- ‘A person can be a person only through other persons’.

To my husband Timothy and both our families… your constant support is priceless All glory belongs to God, whose grace made it possible

CONGRESS PRESENTATIONS AND PUBLICATION

Whati LH, Senekal MS, Steyn NP, Nel JH, Lombard C, Norris S. Development of a valid and reliable nutrition knowledge questionnaire for urban South African adolescents of the Birth-to-Twenty study. 5TH INTERNATIONAL CONFERENCE ON NUTRITION AND FITNESS CONGRESS. JUNE 9-12, 2004, WAR MUSEUM OF ATHENS- HELLAS (GREECE).

Whati LH, Senekal MS, Steyn NP, Nel JH, Lombard C, Norris S. Development of a valid and reliable nutrition knowledge questionnaire for urban South African adolescents of the Birth-to-Twenty study. NUTRITION CONGRESS, AUGUST 23-27, 2004, ATKV GOUDINI SPA HOLIADAY RESORT, WORCESTER. WESTERN CAPE, SOUTH AFRICA

Whati LH, Senekal MS, Steyn NP, Nel JH, Lombard C, Norris S. Development of a valid and reliable nutrition knowledge questionnaire for urban South African adolescents. 2005, Nutrition; 21: 76-85

TABLE

OF

CONTENT

PAGE

CHAPTER 1: INTRODUCTION

1.1 Problem identification and motivation for the study ...2

1.2 Aims ...5

1.3 Objectives ...5

1.4 Descriptions of key concepts ...6

1.5 Outline of the thesis ...6

1.6 References...7

CHAPTER 2: LITERATURE REVIEW

1.1 Overview ...10

1.2 Physical development ...10

1.3 Cognitive development ...12

1.4 Social development ...13

1.5 Implications for nutrition knowledge questionnaire development ...13

2. Adolescence and nutrition ...14

2.1 Nutritional needs of adolescents ...14

2.2 Eating and associated behaviours of adolescents...14

2.3 Factors influencing adolescent eating behaviour ...16

2.3.1 Individual ...16

2.3.2 Social environment ...18

2.3.3 Physical environment...19

2.3.4 Macrosystem ...19

2.4 Nutrition-related health problems experienced in adolescence ...20

2.4.1 Undernutrition...20

2.4.2 Chronic diseases of lifestyle ...21

2.4.3 Other nutrition-related conditions...22

2.5 Implications for nutrition knowledge questionnaire development ...23

3. Nutrition knowledge and adolescence ...24

3.1 Nutrition education for adolescents ...24

3.3 Implications for nutrition knowledge questionnaire development ...31

4. The theory of nutrition knowledge questionnaire development ...31

4.1 Development of a conceptual framework and identification of nutrition concepts...31

4.2 Development of questionnaire item pool ...32

4.3 Construction of preliminary questionnaire ...33

4.4 Refining of preliminary questionnaire through item analysis...33

4.5 Validity and reliability assessment ...34

4.5.1 Validity assessment...34

4.5.2 Reliability assessment...35

4.6 Performance-rating scales ...37

4.6.1 Norm-referenced performance-rating scale ...38

5. Implications for nutrition knowledge questionnaire development ...40

6. References...40

CHAPTER 3: ARTICLE 1

The questionnaire development process ...54

Development of a conceptual framework and identification of concepts...56

Development of questionnaire items...60

Construction of the preliminary questionnaire...62

Refining of the preliminary questionnaire through item analysis and construct validity assessment ...63

Study population (Sample 1)...63

Data collection and processing ...64

Statistical analysis...64

Results and discussion ...66

Assessment of internal consistency reliability of the 63-item questionnaire ...68

Study sample (Sample 2) ...68

Data collection and processing ...68

Statistical analysis...69

Results and discussion ...69

Reassessment of the construct validity and internal consistency reliability of the final

questionnaire ...73

Procedures, methods and data analysis ...73

Results and discussion ...73

General discussion and conclusion ...76

References...78

CHAPTER 4: ARTICLE 2

Development of norms...84

Study design...84

Study population and sampling (Norm group) ...84

Data collection and processing ...85

Data analysis ...85

Results and discussion ...87

Assessing the validity of the norms ...94

Study design...94

Study population (Validation group) ...94

Data collection, processing and analysis...95

Results and discussion ...95

General discussion, conclusion and recommendations...97

References ...98

CHAPTER 5: SUMMARY, CONCLUSION AND RECOMMENDATIONS

2. Final conclusion and recommendations...105

ADDENDUM A:

Final suggested concepts and specific objectives forreview by expert panel

...

ADDENDUM B:

88-item version of the nutrition knowledge questionnaire

...

ADDENDUM C:

Information sheet sent to high school officials and consent forms for Sample 1

...

ADDENDUM D:

Cronbach alpha’s for sections an per section for each item if deleted

...

ADDENDUM E:

Items deleted from 88-item questionnaire after item analysis and rationale for deletion

...

ADDENDUM F:

63-item version of the nutrition knowledge questionnaire

...

ADDENDUM G:

Information sheet sent to high school officials and consent forms for Sample 2

...

ADDENDUM H:

Final 60-item nutrition knowledge questionnaire

...

ADDENDUM I:

Categorisation of items in questionnaire based on the South African Food-Based Dietary Guidelines

...

LIST

OF

TABLES PAGE

guidelines

29

Table 3 Z-scores cut-off points for each stanine in the development of

norms

39

Article 1 Table 1: Final nutrition knowledge concepts of number of items developed

for each concept

60

Table 2: Mean score ±SD and other locations of expert and non-expert

groups (Sample 1) on the 63-item questionnaire

66

Table 3: Cronbach α values of 63-item questionnaire for the study

population (Sample 2) grouped by schools, races and grade levels

70

Table 4: Mean score ±SD and other locations of expert and non-expert

groups (Sample 1) on the 60-item questionnaire

74

Table 5 Cronbach α values of 60-item questionnaire for the study

population (Sample 2) grouped by schools, races and grade levels

75

Article 2 Table 1: Z-score cut-off points for stanines 86

Table 2: Norm-referenced performance-rating scale for the nutrition

knowledge questionnaire developed by Whati et al. 2005

92

Table 3: Column percentages for performance-rating in stanines by

validation group

LIST OF FIGURES

PAGE

knowledge questionnaire for urban adolescents

55

Figure 2 Conceptual framework of nutrition-related issues and causal

factors for urban South African adolescents

58

Figure 3 Race distribution of the study population (Sample 2) 69 Figure 4 Cronbach α values of 63-item questionnaire for the two

schools (Sample 2) according to grades

70

Figure 5 Cronbach α values of 63-item questionnaire for the three

grades (Sample 2) according to schools

71

Figure 6 Cronbach α values of final 60-item questionnaire for the two

schools (Sample 2) according to grades

75

Figure 7 Cronbach α values of final 60-item questionnaire for all grades

(Sample 2) according to schools

76

Article 2 Figure 1 Application of stanines in performance-rating 86 Figure 2 Stanine distribution based on the z-scores of the norm group 88 Figures

2.1-2.12

Stanine distributions of the z-scores of the different grade levels and school groups

98

Figure 3.1-3.4

CHAPTER 1

1.1 PROBLEM IDENTIFICATION AND MOTIVATION FOR THE STUDY

The release of Nelson Mandela from prison in 1990 initiated South Africa’s transformation from an oppressive government serving the interest of a minority to a democratically elected government serving the interests of all citizens. This and other related changes in the country’s political situation had environmental, economical and social implications for the people of South Africa. For example, the formal abolishing of the apartheid laws in 1991 legalised and thereby increased the migration of people from urban to rural areas in their search for jobs, access to educational facilities and for better living conditions (Heribert & Moodley 1993). The Birth-to-Ten study was initiated in 1990 with the aim of investigating biological, environmental, economic and psychosocial factors that are associated with the survival and health of children born and living in such urban areas, particularly in South Africa’s largest metropolis Johannesburg-Soweto, for a period of ten years. The researchers expected children growing up in urban South Africa during this period of transformation to be faced with health challenges relating to adverse sexual and reproductive outcomes, non-communicable diseases and micronutrient deficiencies. These problems were expected to arise because of various biological, social, psychosocial and lifestyle risk factors. Therefore, the researchers wanted to investigate the existence of such factors so as to recommend strategies to reduce or eliminate these risks. The information gathered from this study would enable researchers to understand the determinants of child health and development and thereby provide government and other stakeholders with information that they can use to address the important health-related challenges facing the country’s children during this period of transformation (Richter et al. 2004).

The Birth-to-Ten study continued for 12 years, a period during which several indicators of health and development were assessed on approximately 2 700 of the original 3 723 cohort enrolments. The indicators were birth weight, socio-economic status, household composition, migration patterns, anthropometry, dietary intake, dental health, bone mineral density, lipid profiles, blood pressure, smoking habits and psychosocial stress. The Birth-to-Ten study was successful in achieving its aims during the 12-year period and produced several reports that were circulated to all relevant stakeholders. One of the highlights of the study was its influence in the passing of new legislation prohibiting children access to cigarettes. The results from the study proved to parliament that the legislation at that time was making it easy for children to start smoking and was thereby having a negative impact on the children’s health (Richter et al. 2004). Because of the positive impacts the study was having, a decision was made to continue with the study from year 13 until this cohort reaches age 20 years. From that

point onwards the study was referred to as the Birth-to-Twenty (BTT) study.

Cigarette smoking is only one of the many lifestyle practices that presented a health challenge facing the country. Because of the increased migration from rural to urban areas, the diets of many South Africans was observed to be changing from the traditional high fibre and low fat diet to the western low fibre and high fat and sugar diet, a phenomenon referred to as the nutrition transition. This was also accompanied by an increase in alcohol intake and a decrease in physical activity (Bourne & Steyn 2000; Popkin 2001). This kind of lifestyle has been implicated in the increased prevalence of chronic diseases of lifestyle such as coronary heart disease, type 2 diabetes, hypertension and obesity in South Africa (Walker 2001).

In addition to the South African studies on risk factors and the presence of chronic diseases of lifestyle in adults, studies have also indicated that the risk for certain chronic diseases may appear during childhood and adolescence (Berenson et al. 1998; Selvan & Kurad 2004); Obesity is already a problem among South African adolescents (Puoane et al. 2002; SADHS 1998) and they are also at risk of developing other nutrition-related problems, e.g. eating disorders, and micronutrient deficiencies (Vorster 2002). The possibility of these health challenges presenting in the BTT group and the need for investigations into these potential problems tied in well with the aim and objectives of the study and was adequate motivation for the inclusion of an intensive nutritional assessment component in the BTT study from year 13.

The objectives of this nutritional assessment component in the BTT study were to determine the presence of nutrition–related factors and their association with the eating behaviour of the BTT cohort. The specific nutrition-related factors targeted for investigation included weight management practices, dietary intake, eating attitudes and nutrition knowledge (Richter et al. 1999). These factors can impact on nutrition behaviour and ultimately, individual growth and development. For example, the lack of nutrition knowledge regarding the nutritional value of many commonly consumed foods may lead adolescents to consume foods of poor nutritional value, which may result in nutrient deficiencies, obesity and other chronic diseases of lifestyle (Mahan & Escott-Stump 2004).

Nutrition knowledge assessment was therefore one of the new aspects introduced into the BTT study from age 13 of the cohort. The results will be used in determining whether nutrition knowledge is an independent risk factor and/or linked to other risk factors in the development of nutrition-related diseases in adolescents. For these purposes the BTT research group identified the need for a valid and

reliable nutrition knowledge questionnaire for urban South African adolescents. A literature search showed that such a questionnaire was not available for use on the BTT cohort, indicating the need for the development of an instrument of this nature.

When considering the development of a nutrition knowledge questionnaire for a target group, in this case, the BTT group, the target group needs to be considered. The following was relevant in this regard: the cohort included children residing in the greater Johannesburg metropolitan and surrounding areas whose racial composition included blacks (78%), those of mixed ancestry (12%), whites (6%) and Indians (4%) (Richter et al. 2004). This ‘rainbow nation’ encompassed a variety of languages, cultures, beliefs, attitudes and practices that could impact on nutrition knowledge assessment. A nutrition knowledge questionnaire that would be appropriate for use on the cohort would therefore have to cover all relevant nutrition topics, be in an acceptable and understandable language, be relatively easy to complete and finally, be valid and reliable. Furthermore, the questionnaire should be able to facilitate rating of knowledge levels as well as monitoring possible changes in nutrition knowledge levels (Hopman et al. 2000).

The first step in the development of a questionnaire is to define what level of nutrition knowledge is required from the target group. This is referred to as establishing the measurable attribute (construct) of an instrument (Hawkes & Novak 1998). To accurately define knowledge requires the formulation of a conceptual framework and thereafter identifying nutrition topics (concepts) relevant to the conceptual framework in consultation with an expert panel (Rosander & Sims 1981; Fanslow et al. 1981; Sullivan & Schwartz 1981; Byrd-Bredbenner 1981; Lackey et al. 1981; Stevens et al. 1999; Talip et al. 2003). Following the definition of the conceptual framework and the identification of relevant concepts, an item pool for potential inclusion in the questionnaire must be developed. The items should then be subjected to various evaluations to determine whether they should be included in the final test or not. These evaluations should include assessment by expert panels and pilot testing using a group representative of the target group to ensure content and face validity (Fanslow et al. 1981; Sapp & Jensen.1997; Hawkes & Novak.1998). For statistical assessments such as item analysis for the determination of item difficulty and discrimination, as well as Cronbach  to assess internal consistency (reliability), the questionnaire should be completed by different groups appropriate for such purposes (Parmenter & Wardle 1999). As an assessment of construct validity, the questionnaire should be able to distinguish between groups known to differ in terms of the knowledge being measured (Sapp & Jensen 1997; Hawkes & Nowak 1998).

An important step in ensuring appropriate nutrition knowledge assessment is to develop a performance-rating scale to determine the knowledge levels of a target group. The literature distinguishes between three types of performance-rating scales namely; norm-referenced, criterion referenced, and content-referenced performance-rating scales (Morganthau 1990; McCloskey 1990). A norm-content-referenced performance-rating scale rates a child’s performance in relation to that of a set of scores (norms) that have been established using a standardisation group. The norms can be interpreted according to standard scores, T scores, percentile ranks or stanines. A criterion-referenced scale rates performance against a predetermined standard, which is usually the mastery level, thus comparing the children against each other. These are usually interpreted within an age/gender-related framework. Content-referenced performance-rating scales are based on a range of developmental objectives and are usually interpreted according to the number of objectives that have been accomplished (Morganthau 1990; McCloskey 1990; Taylor & Walton 2001). Finally, the validity of a performance-rating scale should ideally, also be assessed.

It is clear that the development of a valid and reliable nutrition knowledge questionnaire and associated norms should occur through following a suitable well-defined process to ensure that the result are tools that are effective and efficient. Such were the processes that were followed for this study

1.2 AIMS

1.2.1 The primary aim of this study was to develop a valid and reliable nutrition knowledge questionnaire for urban adolescents to be used on participants of the BTT study.

1.2.2. The secondary aim was to develop a performance-rating scale to appropriately and accurately rate the nutrition knowledge of the BTT cohort.

1.3. OBJECTIVES

The following objectives were formulated to realise these aims:

1.3.1 To formulate a conceptual framework for the nutrition knowledge questionnaire and thereby identify the relevant nutrition knowledge concepts.

1.3.2 To develop a pool of potential questions based on the identified nutrition knowledge concepts. 1.3.3 To develop a preliminary questionnaire from the item pool.

1.3.4 To further refine the preliminary questionnaire until acceptable validity and reliability is ensured.

1.3.5 To develop a performance-rating scale to rate nutrition knowledge. 1.3.6 To assess the validity of the developed norms

1.4. DESCRIPTION OF KEY CONCEPTS

1.4.1. Adolescence is a period in a person’s growth and development aimed at ensuring that by the end of that period, he/she has attained the ability to reproduce. This period usually occurs between the ages of 10 and 19 years and involves physical, cognitive and psychosocial growth and development (Dacey & Kenny 1997; United Nations 1997 & Coon 2001).

1.4.2. Nutrition knowledge for urban South African adolescents is defined as knowledge about nutrition that is required for good health as well as for the prevention and management of nutrition– related diseases and conditions, based on current South African recommendations for healthy eating behaviour (see section 3.3).

1.5. OUTLINE OF THE THESIS

After this introduction, the thesis follows the format of four additional chapters. In the literature review (Chapter 2) the different issues that needed to be considered in the development of the questionnaire are discussed in detail. Chapter 3 describes the actual development process that resulted in a 60-item nutrition knowledge questionnaire. This chapter has resulted in the publication of an article in the January 2005 edition of the journal, ‘Nutrition’ (Whati et al. 2005). The second article that is yet to be published is presented in Chapter 4 and describes the process followed to develop a performance-rating scale to rate the nutrition knowledge of testees. In Chapter 5 a summary is provide of the whole process as well as final recommendations regarding the use of the questionnaire and its norms. Lists of tables and figures as well as addenda relevant to the chapters are included in the content pages.

Since the first article has already been published in the journal Nutrition, the referencing system applied throughout this thesis will follow the author’s guidelines of Nutrition. However, the references in the text are not indicated by numbers as prescribed by the journal, but as specified by the Harvard system. Authors are listed in alphabetical order in the reference list.

1.6 REFERENCES

Berenson GS, Srinivasan SR, Nicklas TA. Atherosclerosis: A nutritional disease of childhood. Am J Cardiol 1998; 82(10B): T22-29.

Bourne LT, Steyn K. Rural urban nutrition-related differential among adult population groups in South Africa, with special emphasis on the black population. S Afr J Clin Nutr 2000; 13: S23.

Bryd-Bredbenner C. Nutrition knowledge test for nutrition educators. J Nutr Ed 1981; 113(3): 97-99. Fanslow AM, Brun JK, Hausafus C. The NATs-nutrition achievement tests for the elementary grades. J Nutr Ed 1981; 13: 90-92.

Hawkes A, Novak N. A validated nutrition knowledge questionnaire for cardiac patients. Austr J Nutr Diet 1998; 55(1): 21-24.

Heribert A, Moodley K. The opening of the apartheid mind: Options for the new South Africa. Berkeley: University of Carlifonia Press; 1993. Available at:

http://ark.cdlib.org.ark:/13030/ft958009mm/ Accessed January 3, 2005.

Hopman WM, Towheed T, Anastassiades T, Tenehouse A, Poliquin S, Berger C, et al. Canadian normative data for the SF-36 health survey. Can Med Assoc J 2000; 163(3): 265.

Lackey CJ, Kolasa KM, Penner KP, Mutch BL. Development of the NKT– a general nutrition knowledge test for teachers. J Nutr Ed 1981; 13(3): 100-101.

MacCloskey G. Selecting and using early childhood rating scales. Topics in Early Childhood Special Education. 1990, 10(3)

Mahan LK, Escott-Stump S. Krause’s food, nutrition, and diet Therapy, 11th _{Edition. Pennsylvania:} W.B. Saunders Company; 2004, p. 257-269.

Morganthau T. A consumer guide to testing. Newsweek 1990; 116(10).

Parmenter K, Wardle J. Development of a general nutrition knowledge questionnaire for adults. Eur J Clin Nutr 1999; 53: 298-308.

Puoane T, Steyn K, Bradshaw D, Laubsher R, Fourie J, Lambert V, Mbananga N. Obesity in South Africa: The South African Demographic and Health Survey. Obes Res 2002; 10(10): 1038-1048.

Richter LM, Cameron N, Norris SA, Del Fabro G, MacKeown JM. Birth to Twenty Research Programme Dissemination Report. Johannesburg: University of the Witwatersrand; 2004.

Richter LM, Norris SA, De Wet T. Growing up in the new South Africa “Birth to Ten”- a prospective longitudinal study from birth to 10 years of age. Behav Dev 1999; 36: 5-8.

Rosander K, Sims LS. Measuring effects of an affective – based nutrition education intervention. J Nutr Ed 1981; 13(3): 102-5.

Sapp GS, Jensen HH. Reliability and validity of nutrition knowledge and diet awareness tests developed from 1989 – 1991 diet and health nutrition surveys. J Nutr Ed 1997; 28(2): 61-71.

Selvan MS, Kurpad AV. Primary prevention: Why focus on children and young adolescents. Indian J Med Res 2004; 120: 511-518.

Stevens J, Cornell CE, Story M, French SA, Levin S, Becenti A, Gittelsohn J, Going SB, Reid R. Development of a questionnaire to assess knowledge, attitudes, and behaviour in American Indian adolescents. A J Clin Nutr 1999; 69(4): 773S -781S.

South African Demographic Health Survey (SADHS). Pretoria: Department of Health; 1998.

Sullivan AD, Schwartz N. Assessment of attitudes and knowledge about heart disease. J Nutr Ed 1981; 13(3): 106-108.

Talip W, Steyn NP, Visser M, Charlton KE, Temple M. Development and validation of a knowledge test for health professionals regarding lifestyle modification. Nutrition 2003; 19: 760-766.

Taylor K, Walton S. Who is Norm? And what’s he doing in my class. Instructor 2001; 110(6): 18-19. Vorster HH. The emergence of cardiovascular disease during urbanisation of Africans. Pub Health Nutr 2002; 5(1A): 239-43.

CHAPTER 2

When a nutrition knowledge questionnaire is developed for a particular target group, the nutrition-related characteristics of the target group, general principles and guidelines for good nutrition and the theory for questionnaire development need to be considered. As the questionnaire in this study is aimed at adolescents, the adolescence developmental phase is firstly discussed. This is followed by a discussion on the role of nutrition during adolescence which focuses on the actual nutrient needs of adolescents, their eating behaviour and factors influencing these eating behaviours. The potential health consequences of these behaviours are also discussed. As nutrition knowledge is one of the factors influencing the nutritional behaviour of adolescents and since it is the focus of this study, the third discussion looks at this concept in further detail. Since recommendations for healthy eating behaviour emerged as key concepts in nutrition education and knowledge assessment, these concepts are discussed in more depth. Finally, the theory of nutrition knowledge questionnaire development as well as the development of performance-rating scales is discussed.

1. THE ADOLESCENCE DEVELOPMENTAL STAGE

1.1 Overview

Adolescence is a period during which a child acquires the physical growth and maturity, intellectual/cognitive ability, as well as social knowledge and skills required to live and function as an adult. There are two phases of physical development in the adolescent stage, pubescence and puberty. Pubescence is a stage of physiological growth where primary sex organs mature and various processes that promote reproductive function start to occur. Puberty is the stage during which the processes leading to adult maturity in terms of sexual, emotional, intellectual and physical growth and development begin (Steinberg 1995; Dacey & Kenny 1997; Coon 2001).

1.2 Physical development

Physical development during puberty occurs with the main purpose being for the individual to become capable of reproduction (Rolfes et al. 1998; Coon 2001). In girls, the increased secretion of luitenising hormone (LH) and follicle-stimulating hormone (FSH) by the body’s endocrine system leads to the appearance of secondary sexual characteristics such as breasts and body hair as well as the beginning of menstruation (menarche) which in fertile individuals involves the production and release of ova (ovulation). In boys, the hormone released is testosterone and it promotes the growth of genitalia, appearance of facial and body hair, deepening of the voice, as well as initiation of spermatogenesis

(Dacey & Kenny 1997; Kaplan 2000). During puberty, there is also a rapid increase in height and weight accompanied by changes in regional distribution of body fat resulting in the typical android and gynoid distribution found in older males and females, which is referred to as the ‘growth spurt’ (Dacey & Kenny 1997; Rogol et al. 2002).

There are gender and/or individual differences concerning the onset of puberty, timing of the first menstruation (menarche) in girls, as well as the rate and extent of pubertal growth. Genetics and nutrition have been implicated as the cause of these differences. Genetics has led to the generally earlier onset of puberty in girls of 10-11 years compared to 12-13 years in boys (Dacey & Kenny 1997; Coon 2001). Physically this has no negative consequences since the later onset of puberty in boys implies that they have a longer period of childhood growth and maturity and will ultimately be taller and heavier than girls by the end of puberty (Dacey & Kenny 1997; Rogol et al. 2002). However, the common differences in the onset of puberty between adolescents of the same gender may have a negative or positive impact on the adolescent. The so-called early-maturing or late-maturing adolescents develop at a slower or faster rate than the average adolescent. These non-average adolescents’ body image as well as the society norms (of peers, family, community, etc.) that he/she is exposed to, will determine the type of impact the timing of maturity will have on him/her. For example, boys may react positively to early maturation as it may give them an advantage in the performance of sporting activities while girls may not appreciate looking bigger than their peers if they are treated differently by friends, family and the community or feel awkward for being bigger (Dacey & Kenny 1997; Kaplan 2000; Coon 2001).

Nutrition has an impact on some aspects of pubertal development, specifically in relation to body fat composition in girls. Dietary intake has been implicated in the decrease in onset of menarche from ~17 years of age to the current 12.5 years that was observed in the 20th century. This is because the onset of menarche physically enables girls to produce and release ova thereby allowing the ability to conceive. Since menarche and conception can only occur if there are adequate body fat stores, a diet that promotes early fat storage will lead to an early onset of menarche. Conversely, the typical eating habits of females that characterise anorexia nervosa, which is a common disorder in adolescents, leads to low fat stores which may delay the onset of menarche in such affected girls (Dacey & Kenny 1997; Kaplan 2000).

For the reason that differences are observed in adolescents of the same age and similar gender, the physical growth stage of an adolescent can be characterised by applying the sexual maturity rating

(SMR), referred to as Tanner Ratings. This scale is based on the presence of specific secondary sex characteristics. The ratings range from stage 1 to 5, with 1 indicating the presence of pubescence stage characteristics and 5 indicating the presence of adult characteristics (Dacey & Kenny 1997; Mahan & Escott-Stump 2004).

1.3 Cognitive development

The second aspect of adolescent development involves cognitive development, which relates to the acquisition of skills and abilities that will culminate in the adolescent displaying patterns of thinking that are more advanced, more efficient and more complex than those of younger children thereby reflecting mental maturity (Steinberg 1995; Dacey & Kenny 1997; Kaplan 2000). Cognitive development starts from birth where mental function has not yet become possible and becomes increasingly advanced during early and late childhood. Adolescence is the next developmental stage where cognitive development advances further in that a person learns to formulate their own arguments regarding issues that they previously would respond to by regurgitating the thoughts and opinions of others. This improvement in the manner of thinking is applied in the following four ways: in formulating an individual opinion about how to relate to other people (social cognition); in taking in available information, processing and storing the information in an effective manner (information processing); in logically assessing situations and subsequently making good judgements (critical thinking); as well as in using individual knowledge, skills and imagination to formulate original ideas referred to as creative thinking (Dacey & Kenny, 1997).

In addition to development related to the intellectual functioning of a person, the development of morals is another important aspect of cognitive development. The development of morals refers to the development of the ability to think and reason in a way that leads to behaviour that is acceptable to the individual based on their values and beliefs, as well as to society. The development of morals occurs in three stages. Firstly, thinking is guided by concerns regarding the negative or positive consequences of behaviour (pre-conventional level). Secondly, reasoning is based on following accepted rules and values in wanting to please society (conventional). Finally, self-accepted moral principles that have been carefully conceptualised are followed (post-conventional). Although moral development occurs during adolescents, as part of cognitive development, morals may not have developed to the last stage during adolescence and in fact, continued development or change in morality may occur during adulthood or a person may never reach the conventional or post-conventional stages (Dacey & Kenny 1997; Coon 2001).

1.4 Social development

The third aspect of adolescent development involves social development, which refers to the ways adolescents view themselves as individuals as well as in their interactions and relationships with others. Social development is initiated by the physical and cognitive development that occurs throughout adolescence which leads to adolescents questioning their self as well as their role in society. Self-questioning ultimately leads to a search for identity as well as seeking behavioural and emotional independence (Dacey & Kenny 1997; Kaplan 2000; Coon 2001).

In their search for identity, adolescents seek to develop their own personal ideology regarding values, beliefs, spirituality, relationships and vocational choice. The search occurs through adolescents questioning their current status and later committing themselves to following a specific plan of action. In seeking behavioural and emotional autonomy, adolescents go through a process where they become less reliant on their parents, form relationships with peers and members of the opposite sex and become increasingly dependent on themselves (Steinberg 1995; Dacey & Kenny 1997; Kaplan 2000; Coon 2001).

However, these processes do not go by smoothly and are usually loaded with confusion and conflict. For example, during the search for identity adolescents tend to rely on each other for support through the process of self-discovery. The increasing reliance on peers can extend to a need for their acceptance resulting in poor self-esteem, succumbing to peer pressures, as well as conflict with parents as a result of unacceptable behaviours (Steinberg 1995; Dacey & Kenny 1997; Kaplan 2000; Coon 2001).

1.5 Implications for nutrition knowledge questionnaire development

Nutrition knowledge of the BTT adolescents will be assessed from when the adolescents are at ‘Tanner stage 1’ up to ‘Tanner stage 5’. This means that these adolescents will be at various times and to varying extents, going through the physical changes described above. The changes in physical development will be accompanied by cognitive and social development. Consequently, adolescents will possess various levels of thinking and reasoning abilities throughout the nutrition knowledge assessment period. Their nutrition knowledge and eating behaviours will also be affected by the stage of physical, cognitive and social development. The questionnaire will need to accommodate all these aspects in terms of item formulation and level of presentation in that the items should be formulated in a manner that appeals to the adolescent way of thinking; the form of questioning used should be

acceptable; and the lowest to the highest possible levels of adolescent intellectual abilities should be accommodated.

2. ADOLESCENCE AND NUTRITION 2.1 Nutritional needs of adolescents

The importance of nutrition throughout the lifecycle is undisputable and this is especially true for the adolescent because of the growth spurt experienced during this phase. A general increase in the requirements of all nutrients is experienced during this phase.

Energy needs are based on individual growth rate and level of activity and protein needs are linked to growth patterns (Mahan & Escott-Stump 2004). Adolescents incorporate more calcium, zinc, iron and magnesium into their bodies than at any other stage of the lifecycle. The increase in calcium requirements is linked to accelerated muscular, skeletal, and endocrine development. The build-up of muscle mass and associated greater blood volume in males as well as blood loss associated with the onset of menses in females increases iron requirements for both genders. Zinc is required for growth and development, which is noticeably increased in adolescents. Lastly, magnesium is also increasingly needed for bone growth and the proper functioning of nerves and muscles (National Diary Council,

http://www.nationaldairycouncil.org 2001 accessed August 22, 2002; Mahan & Escott-Stump 2004). The physical growth experienced during adolescence also influences their vitamin requirements. The increased demands for energy lead to increased requirements for thiamine, riboflavin and niacin. Tissue synthesis and cell growth increases vitamins A, C, E, B6, B12 and folic acid needs, while rapid skeletal growth increases vitamin D requirements. The specific role of other micronutrients such as iodine, phosphorus, copper, cobalt, chromium and fluoride in the adolescence phase has not been as extensively studied but it is a matter of course that they also play an important role (Mahan & Escott-Stump 2004).

2.2 Eating and associated behaviours of adolescents

For the purpose of this review the concept ‘eating behaviour’ refers to aspects such as actual eating patterns; food, drink and nutrient intake; as well as factors influencing such patterns and intakes.

Adolescents have been reported to generally display poor eating behaviour which could contribute to the development of nutrition-related diseases and conditions (Story & Resnick 1986; Story et al. 1998).

Firstly, adolescents have been observed to consume less than the recommended amounts of certain essential nutrients, including energy, protein, iron, calcium, fiber, vitamin A, vitamin C, and riboflavin. These deficiencies reflect an inadequate intake of food sources of these nutrients such as fruits, vegetables and milk as well as a diet lacking in variety (Ahmed et al. 1998; Cavadini et al. 2000; Rolland-Cachera et al. 2000; Chapman 2003). In the United States (US), a survey of US adolescent food intake trends from 1965 to 1996 showed that a lack of variety of fiber-rich foods was the cause of inadequate fiber intake in this group (Cavadini et al. 2000). Studies conducted among South African children and adolescents also reported inadequate intakes of energy, calcium, zinc, iron, potassium, riboflavin, iodine, fiber, vitamins B6, B12 as well as folate (Steyn et al. 1990; Walker 1994; Vorster et al. 1997). Black South African adolescents have specifically been found to follow a diet where refined cereals are the staple food and the main type of starch consumed. The lack of whole grain cereals, legumes, and vegetables in their diet was implicated in the poor fiber intake reported in the group (Steyn et al. 1990; Walker 1994).

In contrast to inadequate intake of certain foods or nutrients, over-consumption has also been observed among adolescents. They tend to consume excessive amounts of high fat foods such as hamburgers, fried foods and pastries, resulting in excessive intakes of saturated fat and cholesterol (Sargent et al. 1994; Cavadini et al. 2000). In a South African study by Steyn et al. (1990) a high intake of meat products which resulted in high saturated fat and cholesterol intake levels was reported for white 10-12 year old adolescents in the Western Cape. Lipids and their various components are not the only nutrients taken in excessive amounts. Refined sugar intake in the form of carbonated drinks, confectionary and table sugar (Lytle et al. 2000; Alexy et al. 2002) is also on the on the increase. A high intake of sugars, sweets, cakes and puddings was also reported in the South African adolescents of the Western Cape (Steyn et al. 1990).

Unhealthy weight management practises such as the use of diet pills, vomiting, fasting, and laxatives have been observed in several adolescent groups, with the prevalence of such practices as high as 43% (n=15 349), 57% (n=869) and 57% (n=599) reported in three separate studies in the United States (Grigg et al. 1996; Thombs et al. 1998 & Lowry et al. 2000). Among black first year female university students in South Africa (20 ± 4 years old, n=180) the prevalence of weight management practices was also found to be high with 42% having partaken in these practices in the five years preceding the study. The specific methods used were similar to those observed in the above-mentioned studies but also

included following unhealthy diets obtained from magazines and the skipping of meals (Senekal et al. 2001).

Alcohol use is highly prevalent among adolescents and was observed to start as early as school grades eleven or age 15 in South African adolescents (Myers & Perry 2002; Madu & Matla 2003). Of major concern is the use of alcohol during pregnancy, as was discovered among adolescents in the US (Kokataili et al. 1994). Of the 117 pregnant adolescents recruited for this particular study, 9% were using alcohol. The nutritional implications of alcohol use are discussed in more detail in Section 3.

2.3 Factors influencing adolescent eating behaviour

To understand factors that influence adolescent eating behaviour, Story (2002) identified four levels of influence, namely individual, social environmental, physical environmental and the macrosystem levels.

2.3.1 Individual factors

Individual factors that influence eating behaviour are psychosocial, biological and lifestyle-related (Neumark-Sztainer et al. 1999). The psychosocial factors influencing eating behaviour include the appeal and taste of food, attitude and beliefs, self-efficacy and level of nutrition knowledge. The taste and appeal of food have been identified by Neumark-Sztainer et al. (1999) as the two most common criteria used by adolescents in making food choices indicating that they would consume healthy foods more often if those foods tasted and looked better. Self-efficacy is a person’s belief that they are capable of adopting a specific behaviour (Mounir et al. 1998) and the lack thereof has been described by Crayton (1994) and Neumark-Sztainer et al. (1999) as the reason behind adolescents choosing unhealthy over healthy behaviour. The common attitudes and beliefs of American adolescents regarding food have led to the adolescents’ general categorization of foods as either ‘junk’ or ‘healthy’ foods (Story & Resnick 1986), concepts which have filtered across to other westernised countries. According to American adolescents junk foods are essential for enjoyment, pleasure and socializing, and the consumption of such foods is considered normal and appropriate for their age group and in line with their expression of independence. On the other hand, consuming healthy food was considered to be “abnormal” and only acceptable when such foods are eaten at home as part of a family meal or as part of a weight control diet (Chapman 1993; Neumark-Sztainer et al. 1999). Nutrition knowledge has also been described as one of the psychosocial risk factors for poor nutritional status in adolescence. For example, Walker and Walker (1993) evaluated the nutrition knowledge pertaining to food

composition of White, those of mixed ancestry, and Indian adolescents in South Africa and the results indicated an unsatisfactory level of knowledge among adolescents in all three groups studied. These authors implicated lack of knowledge as the reason for poor eating habits which contributes to the development of degenerative diseases later in the lives of the population groups studied.

Gender and hunger are the biological factors that influence individual eating behaviour (Neumark- Sztainer et al. 1999). Males have been described as more likely to increase intake of certain foods or nutrients as well as regular exercising so as to increase their muscle size, while their female counterparts are more likely to attempt to restrict intake with the aim of loosing weight (Field 2001; McCabe 2002). Gender was also associated with a higher fat and sugar intake in boys compared to girls in the study conducted by Lien et al. (2002) on fifteen year old (n=613) adolescents. On the subject of hunger, the feeling was mentioned and discussed frequently and extensively in the focus group discussions with US adolescents conducted by Neumark-Sztainer et al. (1999) as an important influential factor.

Lifestyle factors influencing adolescent dietary behaviour include perceived barriers to healthy eating such as time constraints, meal patterns, convenience, weight control and cost (Neumark-Sztainer et al. 1999). Adolescents tend to want to sleep longer in the morning thereby often skipping breakfast. Due to very busy schedules skipping meals throughout the day is also very common. When the adolescent eventually takes time to eat, the meals are often not healthy. It can therefore be said that regular meal patterns and perceived time constraints are issues for such adolescents. To complement their busy lifestyles, adolescents tend to consume foods that require minimum preparation at home or outside the home, with the latter’s selection limited to ready-to–eat foods found in convenience stores, fast food restaurants as well as vending machines. In such instances, convenience is an influencing factor on the eating behaviour of adolescents (Neumark-Sztainer et al. 1999; Story 2002).

Attempting to prevent weight gain or attempting to loose weight influences eating behaviour in that adolescents will decide how and what to eat based on the effect it will have on their weight. Weight control behaviours range from the moderate to the extreme. Those practicing moderate levels of weight control tend to practice healthy behaviour such as a high fruit and vegetable intake or limiting saturated fat intake (Story et al. 1998; Neumark-Sztainer et al. 1999). Those who practice extreme weight loss behaviour on the other hand tend to skip meals, induce vomiting and take laxatives and such adolescents were found to consume less fruits and were more likely to consume high fat foods (Story et

With regard to cost, studies have shown that adolescents will buy foods that they can afford. For instance, consumption of fruits and vegetables by adolescents in high schools increased in cafeterias after they had cut the prices by 50%. However, due to their generally lower prices compared to a healthy meal at a restaurant, fast foods are the regular choice of foods consumed by adolescents (Neumark-Sztainer et al. 1999).

2.3.2 Social environment

The social environment for adolescents refers to the family, peers, and society (Story 2002). Although adolescence is a stage in which adolescents seek independence from parents, the family remains a major influence on adolescent eating behaviour through providing access to food and having an effect on an adolescent’s food attitudes, preferences and values (Dacey & Kenny 1997; Kaplan 2000; Story 2002). Family members may also directly or indirectly educate children and each other about healthy or appropriate eating behaviour (Ivanovic et al. 1997) and this knowledge can on its own influence the adolescent’s eating behaviour.

Society can be described as an organised group of individuals who practice certain characteristic behaviours. Although individuals are born and develop further in their capabilities for independent thought, feeling, and action, all these will be influenced by their interaction with society. In addition, society dictates which behaviours are appropriate or acceptable. For instance, society may determine the consumption of certain foods to be taboo while at the same time promoting the consumption of certain foods. All interactions and experiences associated with society that adolescents experience will have an influence on their dietary behaviour (Gifft et al. 1972; Dacey & Kenny. 1997; Kaplan 2000). Peers are such major influences in adolescent behaviour that the issue of peer pressure and adolescent conformity to such pressure is often considered to be a significant part of adolescence (Dacey & Kenny 1997). Adolescents spend most of their time with their peers and rely on their peer group for emotional support and advice. Seeking peer approval regarding eating behaviour is one of adolescents’ common behaviours (Dacey & Kenny. 1997; Kaplan 2000; Neumark-Sztainer et al. 1999). For example, adolescents may practice certain behaviours such as consumption of ‘junk food’ (Neumark-Sztainer et

al. 1999) based on acceptability of such behaviours within their peer group. They may also practice

other behaviours unbeknown to their peers such as purging, so as to conform to peer pressures to achieve thinness (Dacey & Kenny 1997; Rolfes et al. 1998).

2.3.3 Physical environment

The physical environment refers to all places other than the home, where food for consumption is available to adolescents. It includes schools, fast-food restaurants, vending machines, and workplaces. All these are places where adolescents consume foods depending on their availability and accessibility as food sources, as well as their preference for the types of food available in such places (Neumark- Sztainer et al. 1999).

Schools sometimes provide specific foods as per government nutrition programs such as the Integrated Nutrition Program in South Africa (Department of Health, unpublished document) or in cafeterias in the United States (Wills et al. 2004). South African school pupils from age 6-19 years were also found to purchase foods from street vendors selling their goods at schools (Cress-Williams 2001). All these present sources of food or meals that adolescents have access to and thereby form part of their eating environment.

Fast–food outlets hold a great appeal to adolescents due to their accessibility, acceptability, convenience, foods sold and food prices (Story 2002). Fast food outlets provided 32% of meals away from home for adolescents in the US, with increased consumption during weekends. Vending machines have also increased in visibility and popularity, conveniently located to cater for the snacking trend observed in the general population. Convenience stores represent 28% of non-home and non-school eating occasions for US adolescents. US adolescents are also part of the labour force as they take on part-time/weekend jobs (Story 2002). Fast-food outlets and convenience stores are the main suppliers of jobs for adolescents and the workplace can influence the food choice of adolescents due to the types of discounted and sometimes free foods available to them (Nielsen et al. 2002; Story 2002). Although there is inadequate scientific data on South Africa’s physical environment and how it affects the country’s adolescents, one can assume that urbanisation and westernisation is having a similar impact on our adolescents as has been observed in the US.

2.3.4 Macrosystem

According to Story (2002) components of the macrosystem that influence individual eating behaviour include food production, food distribution systems and the media.

With technological, political and economic changes food production has changed in the past few decades from the hunter-gatherer way of life where the focus was on daily survival to an era where

there are large quantities and varieties of food (Walker unpublished document). In the family, the caregiver is the main decision maker regarding how food is distributed amongst members of the household. On a larger scale food producers distribute their products through various channels and government policy has an impact on which, where and how foods are distributed. Government policies may also promote or restrict the import from and export to other counties. This can ultimately be traced to the economic and political situation within a country and relationships with other countries (Gifft et

al. 1972). These factors will all have an impact on the availability and accessibility of foods for

adolescents (Story 2002).

The mass media refers to means of communication that reach large numbers of people at a time. The main ones are print and audio-visual. Print media include newspapers, magazines and books while audiovisual media include radio, television, the internet and films (Story 2002). Adolescents thus live in a media- saturated environment. Coon et al (2001) summarized the results of international peer reviewed studies conducted on children to determine the relationship between their food intake and television viewing. The authors reported that children exposed to advertising choose advertised food products at a higher rate than those who were not exposed. The same applied to requests for food purchases. Unfortunately a similar in-depth study has not been carried out for adolescents. However, television viewing has been implicated in the onset of obesity in US adolescents through the negative impact time spent watching TV and playing computer games has on physical activity (Gortmaker et al. 1990; NDC 2001).

2.4 Nutrition-related health problems experienced by adolescents

2.4.1 Undernutrition

Deficiencies of certain essential micronutrients are common in developing counties and the most common deficiencies that present in adolescents are those of iron, vitamin A, iodine, calcium, and zinc (Korode 1990; Serra-Majem et al. 2001; Venkaiah et al. 2002; Delisle et al. unpublished document). However, less prominent deficiencies of riboflavin, vitamin B6, E and folate (Korode 1990; Serra-Majem et al. 2001) have also been reported. In the South African population there is a general lack of studies conducted in adolescent groups and most available data is on younger children. The latter has revealed the prevalence of vitamin A, vitamin C, iron, riboflavin and niacin deficiencies among young children in urban, peri-urban and rural areas of South Africa (Coutsoudis et al. 1993; MacKeown et al.

1998; Faber et al. 2001). It can be speculated that these deficiencies may also be prevalent among adolescents

In addition to nutrient deficiencies the effects of insufficient food intake has also been observed in South African children. The national food consumption survey of 1999 has reported stunting in 1 out of every 5 children and underweight in 1 out of 10 South African children aged 1-9 years (Labadarios et

al. 2000). Other smaller studies have revealed a 20% and 31% prevalence of stunting among black

children aged 2-5 years and black and coloured children aged 3-9 years (Popkin et al. 1996; Faber et al. 2001). In the adolescent population, a 19.5% prevalence of stunting was observed in South African adolescents (n=860) aged 10-14 years (Rampele et al. 1995).

2.4.2 Chronic diseases of lifestyle

Chronic diseases of lifestyle include conditions such as cardiovascular diseases, cancer, osteoporosis, overweight and obesity as well as Type 2 diabetes that occur in adolescence or later in life (Lytle 2002). Generally, these conditions develop due to a combination of genetic factors and lifestyle patterns such as poor dietary intake, inactivity and cigarette smoking. Therefore one would not expect a high prevalence of such conditions in adolescents. However, overweight and obesity, type 2 diabetes, bone fractures and weight-related eating disorders have already been reported in adolescent populations globally (Lytle & Kubik 2003). Wang et al. (2002) compared the adolescents’ overweight trends in four countries, namely Brazil, Russia, China and the US, spanning three decades (1975- 1998). The prevalence of overweight was found to have increased in three of the four countries studied, namely Brazil (tripled), United States (doubled), and China (increased by one fifth). According to Schneider (2000), this is just an example of the ever increasing prevalence of obesity worldwide.

In South Africa a nationwide study (South African Demographic and Health Survey-SADHS) found 19.2% of South Africans aged 15 and older to be obese (SADHS 1998). A study conducted on 231 first year university students also found 8% of them to be overweight and 6.5% obese (Steyn et al. 2000). Hypertension prevalence increases with age (Health Systems Trust 2000) and the SADHS study revealed that 16% of adult women and 13% of adult men were hypertensive (SADHS 1998). One can thus postulate that the elevated blood pressure reported in the younger populations can develop into hypertension with increasing age.

2.4.3 Other nutrition-related conditions

Eating disorders occur commonly in adolescence and include anorexia nervosa, bulimia nervosa, binge-eating disorders, and anorexic/bulimic behaviours (Rolfes et al. 1998; Lytle & Kubik 2003). In South Africa, one study conducted on first year university students revealed a high prevalence of eating disorders in females, ranging from a 6% prevalence of bulimia/binge-eating to 16% responding in a manner suggesting disturbed eating habits (Sheward 1995). A study which showed a high prevalence of abnormal eating attitudes among white (33%), mixed race (26%) and black (20%) South African adolescents also suggest the presence of eating disorders amongst the group (Caradas et al. 2001). These conditions are psychological disorders that require specialised treatment (Lytle & Kubik 2003) by a team that includes doctors, nurses, psychiatrists and dieticians. Regardless of the aetiology of eating disorders, poor eating habits such as self-starvation, binge eating and purging characterise these conditions (Lytle & Kubik 2003) and they can also lead to the development of chronic illness amongst adolescent girls (Fischer et al. 1995).

HIV/AIDS is a world-wide phenomenon that has reached epidemic proportions in South Africa, with a prevalence of HIV infection of 15.6% reported for South Africans aged 15-49 years of age in 2002 (SAHR 2000). Various opportunistic infections in HIV/AIDS result in disease conditions that are characterised by a decrease in appetite; poor food intake; inefficient metabolism and malabsorption of nutrients; excessive loss of nutrients; and negative drug-nutrient interactions. As a result, a person with HIV/AIDS is at risk of weight loss and wasting; as well as deficiencies of vitamins A, B1, B2, B6, B12, C, E, folate, selenium, zinc and magnesium (Piwoz & Preble 2000; Whitney et al. 2002) and these are all problems that may occur among adolescents with HIV/ AIDS

As was mentioned in section 2.3, alcohol use and abuse is a problem since excessive intake is known to have a negative impact on health status. Alcohol increases the risk for protein–energy malnutrition; cancers of the oral cavity, pharynx, liver, oesophagus and female breast cancer; and anaemia. Nutritional deficiencies occur as a consequence of decreased appetite and thus inadequate intake, as well as changed absorption, metabolism and excretion of essential nutrients. These include protein, calcium, vitamins A, C, B6, thiamine and riboflavin; as well as malabsorption of thiamine, folate, calcium, phosphorus, zinc and vitamin D deficiencies (Williams 1994; Whitney et al. 2002). The risk of iron overload has also been found to increase with the consumption of more than two units of alcoholic drinks per day (Ioannou et al. 2004) and excessive consumption may even lead to diarrhoea as the alcohol inhibits sodium and water absorption (Bode & Bode 2003). Although the presence of

such health conditions as a direct consequence of alcohol intake have not been researched in South African adolescents, the use of and abuse of alcohol by South African adolescents as reported by Myers and Parry (2002) places them at risk of developing such conditions.

Another important issue for South Africans is food and water safety. In certain urban and rural areas of the country there is a lack of access to safe clean water. People are also at risk of consuming unhygienic foods due to the poor hygiene standards observed at some purchase points such as shops and street vendors. For example South African school pupils from age 6-19 years were found to purchase foods from street vendors selling their goods at schools and taxi ranks (Cress-Williams 2001) Poor personal hygiene practices may also lead to the preparation, storage and consumption of unsafe food and water. As a consequence of such practices, adolescents are at risk of developing food and water-borne illnesses such as diarrhoea and botulism which may sometimes prove to be fatal (Whitney

et al. 2002).

The high prevalence of adolescent pregnancy is a worldwide problem with 25% of woman having a child before age 20 (Senderowitz 1995). The rates are higher for South Africa as the SADHS revealed a 35% prevalence of adolescent pregnancy (SADHS 1998). South African adolescents are also at greater risk of early pregnancy as sexual maturation occurs at an early age (13 years) among some adolescent populations (Buga et al. 1996). Pregnancy is associated with nutrition-related consequences which although not directly related to age, are worsened by existing nutritional issues that feature prominently during adolescence. Adolescents who become pregnant while their own nutritional status is poor, including low nutrient stores, are at increased risk for the development of anaemia caused by vitamin B12, folate and iron deficiencies. They are generally at risk of developing vitamin A, vitamin C, calcium and zinc deficiencies. (Gadowsky et al. 1995; Pena et al. 2003; Delisle et al. unpublished document). Pregnant adolescents have also been found to be at risk of being overweight and developing obesity and other chronic diseases later in life (Mahan & Escott- Stump 2004; Delisle et al. unpublished document). In addition, alcohol use in pregnancy could result in foetal alcohol syndrome (Williams 1994). The high prevalence of pregnancy during this stage makes it an important nutrition-related health issue with nutritional implications for female adolescents that needs to be borne in mind.

2.5 Implications for nutrition knowledge questionnaire development

Growth and development occur at a very rapid rate during adolescence and nutrition during this period takes on a special significance. Gender and individual differences occur in the timing, intensity and